Research map / 3D vision / Face and shape
Deformed Implicit Field: Modeling 3D Shapes With Learned Dense Correspondence
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use DIF-Net instead of the strongest listed baseline | ShapeNet category shape reconstruction · ShapeNet-V2 car, airplane, chair, and table test sets | CD (×1000): car: 0.582 -> 0.404; airplane: 0.288 -> 0.249; chair: 0.785 -> 0.661; table: 1.326 -> 1.036 | Helped |
| Use DIF-Net instead of the strongest listed baseline | ShapeNet category shape reconstruction · ShapeNet-V2 car, airplane, chair, and table test sets | EMD: car: 0.037 -> 0.036; airplane: 0.025 -> 0.024; chair: 0.038 -> 0.038; table: 0.046 -> 0.040 | Helped |
| Use DIF-Net for label transfer | Few-shot 3D shape label transfer · ShapeNet-Part car, airplane, chair, and table | IoU (mean|median): average: 64.4 | 67.2 -> 75.2 | 80.0 | Helped |
| Add deformation-based representation (ablation) | DIF-Net · ShapeNet-V2 chair test | CD (×1000): 0.529 -> 0.661 | Hurt |
| Add deformation-based representation (ablation) | DIF-Net · ShapeNet-V2 car test | CD (×1000): 0.353 -> 0.404 | Hurt |
| Add deformation-based representation (ablation) | DIF-Net · ShapeNet-V2 airplane test | CD (×1000): 0.255 -> 0.249 | No clear effect |
| Add deformation-based representation (ablation) | DIF-Net · ShapeNet-V2 table test | CD (×1000): 0.772 -> 1.036 | Hurt |
| Add normal consistency loss (ablation) | DIF-Net correspondence learning · Chair label transfer | IoU | Helped |
| Add deformation smoothness loss (ablation) | DIF-Net correspondence learning · Chair correspondence and texture transfer | correspondence quality | Helped |
| Add the correction field (ablation) | DIF-Net shape representation · Chair shape reconstruction | F-score | Helped |
| Add the correction field (ablation) | DIF-Net correspondence learning · Chair label transfer | label IoU | Helped |
Closest papers by meaning
- Deep Implicit Templates for 3D Shape Representation (CVPR 2021 · similarity 0.91)
- Reduced Representation of Deformation Fields for Effective Non-rigid Shape Matching (NeurIPS 2022 · similarity 0.90)
- Unsupervised Learning of Dense Shape Correspondence (CVPR 2019 · similarity 0.89)
- Implicit Modeling of Non-rigid Objects with Cross-Category Signals (AAAI 2024 · similarity 0.89)
- 4Deform: Neural Surface Deformation for Robust Shape Interpolation (CVPR 2025 · similarity 0.87)
- DNF: Unconditional 4D Generation with Dictionary-based Neural Fields (CVPR 2025 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.